app.py

import gradio as gr
from PIL import Image
import clipGPT
import vitGPT
import skimage.io as io
import PIL.Image
import difflib
import ViTCoAtt
from build_vocab import Vocabulary

def render_image(image_path_or_url):  
    img = Image.open(io.imread(image_path_or_url))
    img = img.resize((80, 80))  # Adjust size as needed
    buf = io.BytesIO()
    img.save(buf, format='JPEG')
    return buf.getvalue()

# Caption generation functions
def generate_caption_clipgpt(image):
    caption = clipGPT.generate_caption_clipgpt(image)
    return caption

def generate_caption_vitgpt(image):
    caption = vitGPT.generate_caption(image)
    return caption

def generate_caption_vitCoAtt(image):
    caption = ViTCoAtt.CaptionSampler.main(image)
    return caption


with gr.Blocks() as demo:
    

    gr.HTML("<h1 style='text-align: center;'>MedViT: A Vision Transformer-Driven Method for Generating Medical Reports 🏥🤖</h1>")
    gr.HTML("<p style='text-align: center;'>You can generate captions by uploading an X-Ray and selecting a model of your choice below</p>")


    with gr.Row():
        model_choice = gr.Radio(["CLIP-GPT2", "ViT-GPT2", "ViT-CoAttention"], label="Select Model")

        generate_button = gr.Button("Generate Caption") 
    
    caption = gr.Textbox(label="Generated Caption") 


    def predict(img, model_name):
        if model_name == "CLIP-GPT2":
            return generate_caption_clipgpt(img)
        elif model_name == "ViT-GPT2":
            return generate_caption_vitgpt(img)
        elif model_name == "ViT-CoAttention":
            return generate_caption_vitCoAtt(img)
        else:
            return "Caption generation for this model is not yet implemented."     


    # Event handlers
    generate_button.click(predict, [image, model_choice], caption)  # Trigger prediction on button click 


demo.launch()